Dr. Virago has a post up right now about how to prioritize certain article submissions over others, with a great discussion going on. Without hi-jacking the comments, I wanted to chime in, because there there are a lot of factors for how different publications become visible, to whom, when, and to what effect.
First off, here's how I weight article submissions in terms of how I would generally submit them.
1) Keep it for my book. A couple of articles out can be great advance press for a book, but too much of the book copy already available seems to leave publishers lukewarm. I have this only on anecdotal evidence, but I have kept several articles on my desk while I work up the book manuscript. And a book at a decent press (i.e. not an academic vanity press, or one with a reputation on the verge of such a thing) will get tenure at most places, and is the kind of scholarly edifice that seems to define careers.
2) Send it out to a top-tier journal. The best journals are visible to a lot of people quickly. Project MUSE sends out a weekly email announcing newly posted issues, and I for one peruse the Tables of Contents pretty much every week, and print off and browse articles that seem interesting. Plus, now that electronic archives are becoming more and more accessible with less and less work, these articles have a long shelf life of high visibility to many readers. And if the work's good, that's very good. Tenure Committees like these, too.
3) See if a good collection cfp comes around: collections are a crap-shoot, though. Very good collections can have a great upside. Bigger name scholars attract repeated visibility, and the more big name scholars in the collection, the higher the visibility. BUT. These collections seem to be few and far between, and often invited anyway, so random CFPs are unlikely to lead to those kind of field-making collections. Still, at this institution, collection articles and journal articles are valued in the same way. This is not true of all places, which sometimes make a distinction that will privilege the top-tier journals. The other thing about collections is that you don't always know what else surrounds your work, so if you respond to a collection cfp, and are accepted, do be sure to ask to see what else is in it--just as high-visibility scholarship can have an afterglow effect on your work, if your good essay is surrounded by crap, that also reflects on you. Collections also seem to have kind of a slow-burn effect. They take longer to get into print, and take longer to reach readers, but because they do sit in libraries, they can have a long slow burn, and sometimes, artilces in collections will pop up longer than journal articles will. Although the advent of pervasive electronic journal publishing, archiving, and access may change this over time.
4) Minor journals, published proceedings, or most online-only journals: First off, there ARE some really amazing online journals out there. PMC (Postmodern Culture) is huge, and Kairos seems to be a big one in composition studies. But a lot of online journals don't have the reputation of print-first journals, even when the quality of the work is very high. I have sent shorter pieces to online journals, and am happy to have them there. But for a major article and a less-well-known online journal or print journal, these are late options for me. I only submit to published proceedings if I know that the conference-length article is the end of the writing for that argument. My one published-proceedings article and one of my online articles falls into that category. The other two online articles (one co-authored, and actually cited in unexpected places, the other forthcoming) are coming out in expanded form in a collection (full disclosure: the one I'm editing, but with the blessing and encouragement of the publisher, precisely because of the promise of expansion). Another article I have out to a minor journal, but only after it's made the rounds to the major journals who have found not to be sufficiently au-courant. If it doesn't get picked up, I'm shelving it. This is, if you remember, the article I pulled from a collection that was coming out from a suspect publisher.
All of this said, there are all kinds of other considerations, and unexpected (and impossible-to-plan-for) ways that the placement of a publication can have an effect.
-->Certain kinds of articles--high accessibility, easy access, easy-to-read articles--may get cited by students more often, which can have a strange kind of after effect.
-->Certain articles can draw certain kinds of other invitations. Of the three times I've been invited to talk without initiating the contact, two came from journals (one major, and another that used to be major, but isn't publishing electronically), and one from a collection.
-->Your institution may or may not have clear guidelines on how they value certain venues, but these may or may not be consistent for outside reviewers, whose voices really do count for something.
There are a lot of other factors that I am not even thinking of here, which is why good mentoring is important: very little of this is intuitive.